max_tokens not work

Nobody has claimed this yet.

```
for chunk in openai.ChatCompletion.create(
model="Yi-34B-Chat-8bits",
messages=[
{
"role": "system",
"content": prompt
},
{
"role": "user",
"content": message
}
],
stream=True,
max_tokens=15000,
# top_p=top_p,
temperature=0.8
):
if hasattr(chunk.choices[0].delta, "content"):
# assistant_response += chunk.choices[0].delta.content
if chunk.choices[0].delta.content is not None:
assistant_response += str(chunk.choices[0].delta.content)
```

Start by reproducing the provided openai.ChatCompletion.create call with the Yi-34B-Chat-8bits model and max_tokens=15000. Trace how max_tokens is handled for streamed responses and define done as the requested token limit being honored; the issue does not mention a file or test to run.

Written by the indexing model from the issue text.